Cancer Causes Control (2012) 23:865–873 DOI 10.1007/s10552-012-9955-4
ORIGINAL PAPER
Hypomethylation of Alu repetitive elements in esophageal mucosa, and its potential contribution to the epigenetic field for cancerization Yasunori Matsuda • Satoshi Yamashita • Yi-Chia Lee • Tohru Niwa • Takeichi Yoshida • Ken Gyobu • Hiroyasu Igaki • Ryoji Kushima • Shigeru Lee • Ming-Shiang Wu • Harushi Osugi • Shigefumi Suehiro • Toshikazu Ushijima Received: 22 August 2011 / Accepted: 5 April 2012 / Published online: 19 April 2012 Ó Springer Science+Business Media B.V. 2012
Abstract Background Aberrant hypermethylation of specific genes is present in esophageal squamous cell carcinomas (ESCCs). Such hypermethylation is also present in normalappearing esophageal mucosae of ESCC patients and is considered to contribute to the formation of a field for cancerization. On the other hand, the presence of global hypomethylation in ESCCs or in their background esophageal mucosae is unknown.
Electronic supplementary material The online version of this article (doi:10.1007/s10552-012-9955-4) contains supplementary material, which is available to authorized users. Y. Matsuda S. Yamashita T. Niwa T. Yoshida K. Gyobu T. Ushijima (&) Division of Epigenomics, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan e-mail:
[email protected] Y. Matsuda S. Lee H. Osugi Department of Gastroenterological Surgery, Graduate School of Medicine, Osaka City University, Osaka, Japan Y.-C. Lee M.-S. Wu Department of Internal Medicine, Collage of Medicine, National Taiwan University, Taipei, Taiwan H. Igaki Esophageal Surgery Division, National Cancer Center Hospital, Tokyo, Japan
Method We collected 184 samples of esophageal mucosae (95 normal mucosae from healthy subjects, and 89 noncancerous background mucosae from ESCC patients) and 93 samples of ESCCs. Methylation levels of repetitive elements (Alu, LINE1) and cancer/testis antigen genes (NY-ESO-1, MAGE-C1) were measured by bisulfite pyrosequencing and quantitative methylation-specific PCR, respectively. Results Methylation levels of Alu, LINE1, NY-ESO-1, and MAGE-C1 were significantly lower in ESCCs than in their background and normal mucosae. Also, in the background mucosae, a significant decrease of the Alu methylation level compared with the normal mucosae was present. In ESCCs, methylation levels of the two repetitive elements and the two cancer/testis antigen genes were correlated with each other. Conclusion This is the first study to show the presence of global hypomethylation in ESCCs, and even in their noncancerous background mucosae. Alu hypomethylation might reflect the severity of an epigenetic field for cancerization. Keywords Esophageal squamous cell carcinoma Hypomethylation Repetitive element Cancer/testis antigen Epigenetics Abbreviations ALDH2 Acetaldehyde dehydrogenase 2 LINE1 Long interspersed nucleotide element 1 PCR Polymerase chain reaction UICC Union for International Cancer Control
R. Kushima Pathology of Clinical Laboratory Division, National Cancer Center Hospital, Tokyo, Japan
Introduction
S. Suehiro Department of Cardiovascular Surgery, Graduate School of Medicine, Osaka City University, Osaka, Japan
Esophageal squamous cell carcinoma (ESCC) remains the predominant histological type of esophageal cancers in
123
866
Asian countries [1–3]. It is known for its high prevalence of multiple occurrences, including synchronous and metachronous occurrence [4, 5]. As for mechanisms of multiple occurrences, it is believed that genetic/epigenetic alterations accumulate in normal-appearing tissues, forming a field for cancerization [6, 7]. We previously demonstrated that aberrant methylation of promoter CpG islands was present in non-cancerous esophageal mucosae of ESCC patients and that methylation of specific genes was associated with smoking history [8] and cancer risk [9]. Because tumor-suppressor genes, such as CDKN2A, CDH1, FHIT, and RASSF1A, are inactivated by methylation of promoter CpG islands in ESCCs [10–13], it was considered that accumulation of aberrant methylation in esophageal mucosae was involved in the formation of a field for ESCCs. In addition to methylation of promoter CpG islands, global hypomethylation is characteristic of tumor cells [14, 15]. Global hypomethylation involves hypomethylation of normally methylated repetitive elements, such as Alu and LINE1 [16], and normally methylated cancer/testis (CT) antigen genes [17, 18]. The hypomethylation of repetitive elements can serve as a surrogate marker for global DNA hypomethylation [19, 20]. ‘‘Normally methylated’’ genes are physiologically methylated and unexpressed in most tissues of adults, except testicular germ cells. In contrast, they are demethylated and expressed in various types of human cancers [21], and specific ones, such as NY-ESO-1 and MAGE-C1, are expected to be useful as potential targets for cancer immunotherapy [22, 23]. The presence of global hypomethylation is associated with increased rates of chromosome recombination and increased incidence of tumor formation [24–26]. Nevertheless, in ESCCs, the presence of global hypomethylation in cancer cells and its potential involvement in the field for cancerization are unknown. In the current study, we aimed to elucidate whether or not hypomethylation is present in ESCCs, and also in noncancerous background esophageal mucosae. To this end, we accurately quantified methylation levels of two repetitive elements, Alu and LINE1, and two CT antigen genes, NY-ESO-1 and MAGE-C1, in ESCCs, their background mucosae, and normal esophageal mucosae from healthy volunteers.
Cancer Causes Control (2012) 23:865–873
male and 8 female; average age = 61, median age = 60, range = 43–85) before any therapeutic intervention (surgery, chemotherapy, or radiation therapy) were collected by endoscopic biopsy at the National Taiwan University Hospital, Taipei, Taiwan. In 89 of the 93 ESCC patients, we collected biopsy samples not only from cancerous lesions but also from non-cancerous background mucosae (82 male and 7 female; average age = 61, median age = 60, range = 43–85). A total of 277 samples for methylation analysis were collected. Additionally, for immunohistochemical staining, surgical specimens of primary ESCCs and their paired non-cancerous background mucosae were collected from 48 patients (41 male and 7 female; average age = 63, range = 41–83) who underwent esophagectomy without any neoadjuvant therapy at the National Cancer Center Hospital, Tokyo, Japan. Noncancerous background mucosa was defined as the area that was at least 5 cm away from the cancerous margin and appeared normal by iodine staining. All samples were stored at -80 °C after biopsy or resection until the extraction of genomic DNA. Informed consents and interviews for lifestyle risk factors, including cigarette smoking and alcohol intake, were obtained from all the individuals. The range of follow-up period after endoscopic examination was 461–944 days. Disease stages were classified according to the 7th edition of the TNM classification by UICC. Bisulfite pyrosequencing for repetitive elements Sodium bisulfite modification was performed using 1 lg of BamHI-digested genomic DNA as previously described [27]. The modified DNA was suspended in 40 ll of Tris– EDTA buffer, and an aliquot of 1 ll was used for bisulfite pyrosequencing. The CpG sites that showed most significant hypomethylation in gastric cancer [28] were selected for analysis in this study. All primers for pyrosequencing were the same as those we previously reported [28] and are listed in Supplementary Table 1. The PCR products labeled with biotin were annealed to 0.2 lM pyrosequencing primers, and pyrosequencing was carried out using the PSQ 96 Pyrosequencing System (Qiagen, CA, USA). Methylation levels were obtained using PSQ Assay Design software (Qiagen). Quantitative methylation-specific PCR (qMSP)
Materials and methods Patients and tissue samples Normal mucosae of 95 healthy volunteers (69 male and 26 female; average age = 58, median age = 59, range = 25–91) and cancerous lesions of 93 ESCC patients (85
123
Real-time MSP was performed with a primer set specific to methylated (M) or unmethylated (U) sequence by using a 2 ll aliquot of the sodium bisulfite-treated DNA, SYBR Green (Bio Whittaker Molecular Applications, MD, USA), and an iCycler Thermal Cycler (Bio-Rad Laboratories, CA, USA). MSP primers and PCR conditions are shown in Supplementary Table 2. DNA methylated by SssI
Cancer Causes Control (2012) 23:865–873
methylase (New England Biolabs, MA, USA) and DNA amplified twice by a GenomiPhi DNA amplification kit (GE Healthcare Bio-Science, Buckinghamshire, England) were used as fully methylated and unmethylated control DNA, respectively. As in our previous report [8], methylation levels were calculated as the fraction of methylated molecules in the total number of DNA molecules (the number of methylated molecules ? the number of unmethylated molecules). Immunohistochemistry of CT antigens Immunohistochemical staining of NY-ESO-1 and MAGEC1 antigens was performed using a mouse monoclonal antiNY-ESO-1 antibody (clone E978, Invitrogen, CA, USA) and anti-MAGE-C1 antibody (clone CT7-33, DAKO, CA, USA) as primary antibodies. Formalin-fixed and paraffin-embedded samples were sliced at 3 lm thickness, deparaffinized, and heated in 10 mM citrate buffer (pH 6.0) for 20 min (NY-ESO-1) and 5 min (MAGE-C1) at 120 °C by autoclave. After blocking with Blocking-One (Nacalai Tesque, Kyoto, Japan), the sections were incubated with a primary antibody at a concentration of 2.5 lg/ml (NY-ESO-1) or 0.21 lg/ml (MAGE-C1) at 4 °C overnight. Detection of the primary antibody was performed with the DAKO Envision Plus system at room temperature for 60 min and DAB (Wako, Osaka, Japan) as a chromogen. Slides were counterstained with hematoxylin. As a positive control, we used the normal part of a testis with intact spermatogenesis collected from an adolescent patient with seminoma.
867
(42.0 ± 3.8%, mean ± SD) than in non-cancerous background mucosae of cancer patients (45.7 ± 2.3%) and normal mucosae of healthy volunteers (46.5 ± 2.4%) (p \ 0.001) (Fig. 1a). Notably, the mean methylation level of Alu was significantly lower even in the background mucosae than in the normal mucosae (p = 0.018). The mean methylation level of LINE1 was significantly lower in ESCCs (62.2 ± 12.1%) than in the normal and background mucosae (79.7 ± 3.9%, 78.8 ± 6.7%, respectively, p \ 0.001) (Fig. 1b). Unlike Alu, significant differences between the normal and background mucosae were not observed. Methylation levels of CT antigen genes were quantified by real-time MSP (Fig. 1c, d). Both NY-ESO-1 and MAGEC1 were almost fully methylated in the normal and background mucosae but demethylated in 22 (23.7%) and 23 (24.7%) samples of ESCCs, respectively, with a threshold value of 95%. The mean methylation levels of NY-ESO-1 and MAGE-C1 were significantly lower in ESCCs (89.3 ± 22.4%, 94.0 ± 14.9%, respectively) than in their background (99.6 ± 1.4%, 99.8 ± 0.6%, respectively) and normal mucosae (99.7 ± 0.9%, 99.7 ± 2.2%, respectively) (p \ 0.001 each). Correlation among methylation levels of repetitive elements, CT antigen genes, and other promoter CpG islands
Results
We analyzed correlation of methylation levels among two repetitive elements, two CT antigen genes, and four promoter CpG islands in the background mucosae and in ESCCs (Table 1 and Supplementary Fig. 1). As promoter CpG islands hypermethylated in background mucosae, we selected four genes, HOXA9, NEFH, UCHL1, and MT1M, whose methylation levels were analyzed in our previous study (Supplementary Fig. 2) [9]. In the background mucosae, methylation levels were significantly correlated among the four genes. Also, significant negative correlations were observed between the hypomethylation of repetitive elements and hypermethylation of the four genes. In ESCCs, significant positive correlation was observed among individual repetitive elements and CT antigen genes. Correlations between Alu and LINE1 hypomethylation and between NY-ESO-1 and MAGE-C1 hypomethylation were especially stronger than the other correlations.
Hypomethylation of repetitive elements and CT antigen genes in ESCCs, and Alu hypomethylation in background mucosae
The lack of association between hypomethylation and exposure to risk factors, and between hypomethylation and clinicopathological findings
Methylation levels of repetitive elements were quantified by bisulfite pyrosequencing in a total of 277 samples. The mean methylation level of Alu was significantly lower in ESCCs
We first analyzed correlation between the age and degree of hypomethylation of the two repetitive elements. No significant correlation was observed in healthy volunteers
Statistical analysis Differences in mean methylation levels were analyzed by Student’s t test (when variances were equal) and Welch’s t test (when variances were unequal). Correlations of methylation levels among individual repetitive elements and promoter CpG islands were analyzed using Pearson’s product-moment correlation coefficients. All the analyses were performed using PASW statistics version 18.0 (SPSS Japan Inc., Tokyo, Japan), and the results were considered significant when p values\0.05 were obtained by a two-sided test.
123
868
a
Alu 100
Methylation Level (%)
Fig. 1 Methylation levels of the two repetitive elements and the two CT antigen genes in normal mucosae of healthy subjects (n = 95), noncancerous background mucosae of cancer patients (n = 89), and ESCCs (n = 93) collected by endoscopic biopsy. Distribution of the methylation levels at a particular CpG site of Alu (a), LINE1 (b), NY-ESO-1 (c), or MAGE-C1 (d) is shown, and a horizontal line in a chart represents a mean methylation level for each group. The mean methylation levels of Alu, LINE1, NY-ESO-1, and MAGEC1 in ESCCs were significantly lower than those in normal and background mucosae. In addition, Alu methylation levels were significantly lower in the background mucosae than in the normal mucosae
Cancer Causes Control (2012) 23:865–873
b **
*
**
**
* p < 0.05 ** p < 0.001
LINE1 **
90 80 70 60 50 40 30 20 Normal (n = 95)
Background (n = 89)
ESCC (n = 93)
Normal (n = 95)
NY-ESO-1
c
**
Background (n = 89)
** p < 0.001
MAGE-C1
d
**
**
ESCC (n = 93)
**
Methylation Level (%)
100 80 60 40 20 0 Normal (n = 95)
Background (n = 89)
(Alu; r = -0.181, p = 0.080, LINE1; r = 0.129, p = 0.214) or in ESCC patients (Alu; r = -0.085, p = 0.426, LINE1; r = 0.010, p = 0.925), showing the hypomethylation was not age-dependent. Then, in normal mucosae, background mucosae, and ESCCs, we analyzed association between exposure to risk factors and the degree of hypomethylation of the two repetitive elements and the two CT antigen genes. In any group of samples, methylation levels were not associated with history of cigarette smoking or alcohol intake (Table 2). Even when the samples were classified according to the genetic polymorphisms, whether they had an active ALDH2 allele (ALDH21/ALDH21 homozygote) or an inactive ALDH2 allele (ALDH21/ALDH22 heterozygote and ALDH22/ALDH22 homozygote), methylation levels were not associated with alcohol intake (data not shown). In ESCCs, we also analyzed association between methylation levels of repetitive elements or CT antigen genes and clinicopathological findings, including depth of tumor, tumor differentiation, lymph node metastasis,
123
ESCC (n = 93)
Normal (n = 95)
Background (n = 89)
ESCC (n = 93)
multiplicity of tumor, and tumor recurrence (Table 3). There was no significant association between methylation levels of repetitive elements in ESCCs and clinicopathological findings, and between methylation levels of CT antigen genes and clinicopathological findings. Association between hypomethylation and expression of CT antigen genes Finally, we aimed to assess association between hypomethylation and expression of CT antigen genes using immunohistochemistry. From 48 surgical specimens, we selected six ESCCs that had various degrees of hypomethylation and their paired non-cancerous background mucosae (Table 4; representative results in Fig. 2 and Supplementary Fig. 3). For NY-ESO-1, positive staining was observed in two of four ESCCs with demethylation and none of two background mucosae with demethylation. For MAGE-C1, positive staining was observed in three of five ESCCs with demethylation and none of one
Cancer Causes Control (2012) 23:865–873
869
Table 1 Correlation among hypomethylation of individual repetitive elements, CT antigens, and hypermethylation of promoter CpG islands Alu
LINE1
NY-ESO-1
MAGE-C1
HOXA9
NEFH
UCHL1
MT1M
Background mucosae Alu
–
LINE1
0.063
0.152
-0.132
-0.351
-0.240
-0.231
-0.398
–
0.095
0.092
-0.313
-0.464
-0.441
-0.554
NY-ESO-1
–
MAGE-C1
0.038
-0.356
-0.024
0.010
-0.238
–
-0.088
-0.002
0.008
0.037
–
0.341
0.632
0.490
–
0.561
0.544
–
0.312
HOXA9 NEFH UCHL1 MT1M
–
ESCC Alu LINE1
–
0.502 –
0.307 0.470
NY-ESO-1
–
MAGE-C1
0.240 0.383
-0.162 -0.386
0.093 0.189
-0.002 -0.049
0.181 -0.262
0.566
-0.182
0.240
0.139
-0.019
–
-0.268
0.171
0.060
0.082
–
0.057
0.133
0.178
–
0.268
-0.085
–
0.381
HOXA9 NEFH UCHL1 MT1M
–
Coefficient values shown in bold type denote that p value is \0.05
Table 2 Association between hypomethylation and exposure to risk factors Sample
N
Alu
LINE1
Mean ± SD
p value
0.411
NY-ESO-1
Mean ± SD
p value
79.0 ± 3.6
0.054
MAGE-C1
Mean ± SD
p value
99.8 ± 0.4
0.474
Mean ± SD
p value
99.5 ± 2.9
0.316
Normal mucosae Smoking history (-)
53
46.3 ± 2.2
(?)
42
46.7 ± 2.6
80.6 ± 4.2
99.6 ± 1.3
99.9 ± 0.2
Alcohol history (-)
61
46.5 ± 2.3
(?)
34
46.4 ± 2.6
0.837
79.5 ± 3.8
0.517
80.1 ± 4.1
99.8 ± 0.2
0.096
99.4 ± 1.4
99.9 ± 0.2
0.337
99.3 ± 3.6
Background mucosae Smoking history (-)
15
45.1 ± 2.2
(?)
74
45.8 ± 2.3
0.324
79.5 ± 4.9
0.650
78.7 ± 7.0
99.5 ± 0.9
0.786
99.6 ± 1.5
99.9 ± 0.2
0.683
99.8 ± 0.7
Alcohol history (-)
11
45.6 ± 1.0
(?)
78
45.7 ± 2.4
0.866
79.8 ± 5.6
0.613
78.7 ± 6.8
99.5 ± 1.1
0.828
99.6 ± 1.4
99.9 ± 0.2
0.623
99.8 ± 0.7
ESCC Smoking history (-)
17
42.7 ± 3.1
(?)
76
41.9 ± 3.9
0.433
61.3 ± 13.0
0.734
62.4 ± 11.9
77.4 ± 34.1
0.106
91.9 ± 18.1
86.6 ± 26.4
0.184
95.7 ± 10.4
Alcohol history (-)
12
41.7 ± 3.6
(?)
81
42.1 ± 3.8
0.746
61.4 ± 15.9 62.3 ± 11.5
0.810
85.1 ± 33.6 89.9 ± 20.4
0.495
85.8 ± 30.8
0.316
95.2 ± 10.6
123
870
Cancer Causes Control (2012) 23:865–873
Table 3 Association between hypomethylation and clinicopathological findings Variable
N
Alu
LINE1
NY-ESO-1
MAGE-C1
Mean ± SD
p value
Mean ± SD
p value
Mean ± SD
p value
Mean ± SD
p value
0.146
64.9 ± 8.6
0.399
89.3 ± 20.7
0.894
95.2 ± 5.9
0.737
Depth of tumor T1/T2
9
40.2 ± 3.0
T3/T4
66
42.1 ± 3.6
Poorly
12
41.2 ± 5.0
Moderately/Well
33
42.3 ± 4.0
Lymph node metastasis Negative 11
40.7 ± 3.4
61.2 ± 12.5
90.3 ± 21.6
93.3 ± 17.2
Tumor differentiation
Positive
65
0.463
60.1 ± 14.7
0.932
60.5 ± 10.8 0.208
42.1 ± 3.6
92.1 ± 16.6
0.665
88.9 ± 23.1
60.6 ± 8.8
0.715
62.0 ± 12.6
98.0 ± 3.2
0.288
91.7 ± 20.0
95.7 ± 12.7
0.369
89.4 ± 22.3
97.5 ± 3.5
0.392
92.9 ± 17.3
Multiplicity of tumor Solitary
74
41.8 ± 3.9
Multiple
8
42.5 ± 3.8
0.641
61.6 ± 12.3
0.251
62.0 ± 11.0
87.3 ± 24.5
0.240
97.6 ± 5.6
93.0 ± 16.5
0.310
99.0 ± 1.3
Tumor recurrence Negative
79
41.8 ± 3.7
Positive
9
42.8 ± 3.8
Table 4 Association between hypomethylation and expression of CT antigen genes
a
Microscopic findings are shown in Fig. 2 b Microscopic findings are shown in Supplementary Fig. 3
0.460
Sample ID.
62.5 ± 12.5
89.3 ± 22.2
Methylation level (%) Background mucosae ES-29 93.3 ES-39 87.9 ES-49 99.8 ES-58 97.5 ES-59 100.0 ES-62 100.0 ESCC ES-29 12.2 ES-39 80.3 ES-49 51.0 ES-58 53.1 ES-59 99.7 ES-62 99.4
Discussion In this study, we demonstrated that hypomethylation of Alu and LINE1 was present in ESCCs. Although the presence of global hypomethylation is considered to be a universal phenomenon across various types of cancers [14, 29–31], this is the first report on its presence in ESCCs and
0.609
93.2 ± 16.5
NY-ESO-1
background mucosa with demethylation. There was no specimen that stained positive without demethylation, supporting the role of methylation status in expression regulation of the CT antigen genes.
123
0.373
58.6 ± 10.7
93.4 ± 16.0
0.393
98.0 ± 3.3
MAGE-C1 Immunohistochemistry
Methylation level (%)
Immunohistochemistry
-a -
100.0 99.1 99.9 93.7 99.8 100.0
-b -
?a (diffuse) ?a (hetero) -a -
72.1 92.7 51.9 54.0 93.7 97.4
?b (hetero) ? (hetero) ? (hetero) -b
provides a fundamental piece of information. Also, it was striking that hypomethylation of Alu, but not LINE1, was present in non-cancerous background mucosae of ESCC patients. This strongly indicated that Alu hypomethylation is induced earlier than LINE1 hypomethylation during carcinogenesis of ESCCs. The Alu hypomethylation was considered to be involved in the epigenetic field for cancerization, and might have potential as an epigenetic marker for ESCC risk estimation. In the background mucosae, hypomethylation of repetitive elements negatively correlated with hypermethylation of promoter CpG islands of the four genes, HOXA9, NEFH, UCHL1, and MT1M. This finding suggested a possibility
Cancer Causes Control (2012) 23:865–873
871
Fig. 2 Representative immunohistochemical staining of NY-ESO-1 in surgical specimens. a Non-cancerous background mucosa (ES-59, methylation level = 100.0%), b ESCC with full methylation (ES-59, methylation level = 99.7%), c ESCC with partial demethylation (ES58, methylation level = 53.1%), and d ESCC with almost complete demethylation (ES-29, methylation level = 12.2%) are presented. A scale bar represents 100 lm. Neither the background specimen nor
the specimen of ESCC with full methylation had staining. Heterogeneous staining, mainly nuclear, was observed in the specimen with partial demethylation. Diffuse staining in both cytoplasm and nucleus was observed in the specimen with almost complete demethylation. The result of immunohistochemistry was consistent with methylation status of CT antigen genes
that global hypomethylation and hypermethylation of some promoter CpG islands were caused by shared factors. In our study, however, while hypermethylation of these four genes is known to correlate with smoking history [8, 9], no association was found between hypomethylation and exposure to risk factors, including history of cigarette smoking and alcohol intake. It was considered that cigarette smoking and alcohol intake were more closely associated with hypermethylation of these CpG islands than with global hypomethylation. In the stomach, we previously showed that hypomethylation of Alu, LINE1, and Sata is present and that Alu hypomethylation was more sensitive to hypomethylation due to Helicobacter pylori infection than the other repetitive elements [28]. Taken together, Alu seems to be most susceptible to hypomethylation due to exposure to carcinogenic stimuli. The different susceptibility to hypomethylation between Alu and LINE1 might be attributed to the differences in their intrinsic functions. Alu depends on the proteins encoded by LINE1 for its retrotranscription and transposition [32], whereas LINE1 is transposable autonomously [33]. Therefore, there is a possibility that LINE1
hypomethylation, which can induce its retrotranscription and transposition [34], is more strictly regulated than Alu hypomethylation. We also showed that CT antigen genes were demethylated in a coordinated way with global hypomethylation in ESCCs and expression of these genes was consistent with methylation states of each sample by immunohistochemistry. These data suggested that CT antigen genes tended to be demethylated in ESCCs with global hypomethylation. This raises a possibility that induction of CT antigen by DNA methyltransferase inhibitors [35, 36] might lead to a more effective cancer immunotherapy targeting CT antigens. Methylation levels of repetitive elements and CT antigen genes in ESCCs were not associated with any clinicopathological features. To exclude the possibility that a fraction of cancer cells were highly variable among the biopsy samples and made associations undetectable, we microscopically analyzed the fraction of cancer cells in biopsy samples. The fraction was 67.3 ± 6.8% (mean ± SD), and such a possibility seemed to be low. However, there still remains a possibility that use of samples with high purity, such as those prepared by laser-captured microdissection, might reveal
123
872
Cancer Causes Control (2012) 23:865–873
some weak association between hypomethylation and clinicopathological findings. In conclusion, our study demonstrated that hypomethylation of Alu and LINE1 was present in ESCCs and that Alu hypomethylation was present even in non-cancerous background mucosae of ESCC patients. It was suggested that Alu hypomethylation represents the severity of an epigenetic field for cancerization and might become an epigenetic marker for ESCC risk estimation. Acknowledgments This study was supported by Grants-in-Aid for the National Cancer Center Research and Development Fund, and by the Project for Development of Innovative Research on Cancer Therapeutics (P-Direct) from the Ministry of Education, Culture, Science and Sport, Japan. Y. M. and K. G. are recipients of a Research Resident Fellowship from the Foundation for Promotion of Cancer Research. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Shibata A, Matsuda T, Ajiki W, Sobue T (2008) Trend in incidence of adenocarcinoma of the esophagus in Japan, 1993–2001. Jpn J Clin Oncol 38:464–468 2. Lu CL, Lang HC, Luo JC, Liu CC, Lin HC, Chang FY, Lee SD (2009) Increasing trend of the incidence of esophageal squamous cell carcinoma, but not adenocarcinoma, in Taiwan. Cancer Causes Control 21:269–274 3. Hongo M, Nagasaki Y, Shoji T (2009) Epidemiology of esophageal cancer: orient to occident. Effects of chronology, geography and ethnicity. J Gastroenterol Hepatol 24:729–735 4. Pesko P, Rakic S, Milicevic M, Bulajic P, Gerzic Z (1994) Prevalence and clinicopathologic features of multiple squamous cell carcinoma of the esophagus. Cancer 73:2687–2690 5. Shimizu Y, Tukagoshi H, Fujita M, Hosokawa M, Kato M, Asaka M (2001) Metachronous squamous cell carcinoma of the esophagus arising after endoscopic mucosal resection. Gastrointest Endosc 54:190–194 6. Slaughter DP, Southwick HW, Smejkal W (1953) Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer 6:963–968 7. Ushijima T (2007) Epigenetic field for cancerization. J Biochem Mol Biol 40:142–150 8. Oka D, Yamashita S, Tomioka T, Nakanishi Y, Kato H, Kaminishi M, Ushijima T (2009) The presence of aberrant DNA methylation in noncancerous esophageal mucosae in association with smoking history: a target for risk diagnosis and prevention of esophageal cancers. Cancer 115:3412–3426 9. Lee YC, Wang HP, Wang CP, Ko JY, Lee JM, Chiu HM, Lin JT, Yamashita S, Oka D, Watanabe N et al (2011) Revisit of field cancerization in squamous cell carcinoma of upper aero digestive tract: better risk assessment with epigenetic markers. Cancer Prev Res 4:1982–1992 10. Xing EP, Nie Y, Song Y, Yang GY, Cai YC, Wang LD, Yang CS (1999) Mechanisms of inactivation of p14ARF, p15INK4b, and p16INK4a genes in human esophageal squamous cell carcinoma. Clin Cancer Res 5:2704–2713
123
11. Si HX, Tsao SW, Lam KY, Srivastava G, Liu Y, Wong YC, Shen ZY, Cheung AL (2001) E-cadherin expression is commonly downregulated by CpG island hypermethylation in esophageal carcinoma cells. Cancer Lett 173:71–78 12. Tanaka H, Shimada Y, Harada H, Shinoda M, Hatooka S, Imamura M, Ishizaki K (1998) Methylation of the 50 CpG island of the FHIT gene is closely associated with transcriptional inactivation in esophageal squamous cell carcinomas. Cancer Res 58:3429–3434 13. Wong ML, Tao Q, Fu L, Wong KY, Qiu GH, Law FB, Tin PC, Cheung WL, Lee PY, Tang JC et al (2006) Aberrant promoter hypermethylation and silencing of the critical 3p21 tumour suppressor gene, RASSF1A, in Chinese oesophageal squamous cell carcinoma. Int J Oncol 28:767–773 14. Feinberg AP, Vogelstein B (1983) Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature 301:89–92 15. Jones PA, Baylin SB (2007) The epigenomics of cancer. Cell 128:683–692 16. Ehrlich M (2002) DNA methylation in cancer: too much, but also too little. Oncogene 21:5400–5413 17. De Smet C, De Backer O, Faraoni I, Lurquin C, Brasseur F, Boon T (1996) The activation of human gene MAGE-1 in tumor cells is correlated with genome-wide demethylation. Proc Natl Acad Sci USA 93:7149–7153 18. Kaneda A, Tsukamoto T, Takamura-Enya T, Watanabe N, Kaminishi M, Sugimura T, Tatematsu M, Ushijima T (2004) Frequent hypomethylation in multiple promoter CpG islands is associated with global hypomethylation, but not with frequent promoter hypermethylation. Cancer Sci 95:58–64 19. Yang AS, Estecio MR, Doshi K, Kondo Y, Tajara EH, Issa JP (2004) A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res 32:e38 20. Weisenberger DJ, Campan M, Long TI, Kim M, Woods C, Fiala E, Ehrlich M, Laird PW (2005) Analysis of repetitive element DNA methylation by MethyLight. Nucleic Acids Res 33:6823– 6836 21. Caballero OL, Chen YT (2009) Cancer/testis (CT) antigens: potential targets for immunotherapy. Cancer Sci 100:2014–2021 22. Gnjatic S, Nishikawa H, Jungbluth AA, Gure AO, Ritter G, Jager E, Knuth A, Chen YT, Old LJ (2006) NY-ESO-1: review of an immunogenic tumor antigen. Adv Cancer Res 95:1–30 23. Nuber N, Curioni-Fontecedro A, Matter C, Soldini D, Tiercy JM, von Boehmer L, Moch H, Dummer R, Knuth A, van den Broek M (2010) Fine analysis of spontaneous MAGE-C1/CT7-specific immunity in melanoma patients. Proc Natl Acad Sci USA 107:15187–15192 24. Chen RZ, Pettersson U, Beard C, Jackson-Grusby L, Jaenisch R (1998) DNA hypomethylation leads to elevated mutation rates. Nature 395:89–93 25. Eden A, Gaudet F, Waghmare A, Jaenisch R (2003) Chromosomal instability and tumors promoted by DNA hypomethylation. Science 300:455 26. Gaudet F, Hodgson JG, Eden A, Jackson-Grusby L, Dausman J, Gray JW, Leonhardt H, Jaenisch R (2003) Induction of tumors in mice by genomic hypomethylation. Science 300:489–492 27. Kaneda A, Kaminishi M, Sugimura T, Ushijima T (2004) Decreased expression of the seven ARP2/3 complex genes in human gastric cancers. Cancer Lett 212:203–210 28. Yoshida T, Yamashita S, Takamura-Enya T, Niwa T, Ando T, Enomoto S, Maekita T, Nakazawa K, Tatematsu M, Ichinose M et al (2011) Alu and Satalpha hypomethylation in Helicobacter pylori-infected gastric mucosae. Int J Cancer 128:33–39 29. Wilson AS, Power BE, Molloy PL (2007) DNA hypomethylation and human diseases. Biochim Biophys Acta 1775:138–162
Cancer Causes Control (2012) 23:865–873 30. Kim BH, Cho NY, Shin SH, Kwon HJ, Jang JJ, Kang GH (2009) CpG island hypermethylation and repetitive DNA hypomethylation in premalignant lesion of extrahepatic cholangiocarcinoma. Virchows Arch 455:343–351 31. Ogino S, Nosho K, Kirkner GJ, Kawasaki T, Chan AT, Schernhammer ES, Giovannucci EL, Fuchs CS (2008) A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer. J Natl Cancer Inst 100:1734–1738 32. Dewannieux M, Esnault C, Heidmann T (2003) LINE-mediated retrotransposition of marked Alu sequences. Nat Genet 35:41–48 33. Belancio VP, Hedges DJ, Deininger P (2008) Mammalian nonLTR retrotransposons: for better or worse, in sickness and in health. Genome Res 18:343–358
873 34. Bourc’his D, Bestor TH (2004) Meiotic catastrophe and retrotransposon reactivation in male germ cells lacking Dnmt3L. Nature 431:96–99 35. Guo ZS, Hong JA, Irvine KR, Chen GA, Spiess PJ, Liu Y, Zeng G, Wunderlich JR, Nguyen DM, Restifo NP et al (2006) De novo induction of a cancer/testis antigen by 5-aza-20 -deoxycytidine augments adoptive immunotherapy in a murine tumor model. Cancer Res 66:1105–1113 36. Oi S, Natsume A, Ito M, Kondo Y, Shimato S, Maeda Y, Saito K, Wakabayashi T (2009) Synergistic induction of NY-ESO-1 antigen expression by a novel histone deacetylase inhibitor, valproic acid, with 5-aza-20 -deoxycytidine in glioma cells. J Neurooncol 92:15–22
123